In a communications network, in order to improve reliability of network communications or increase a processing capability of a network node, multiple communication nodes are usually deployed on network planes at a same level on a communication path. After one of the communication nodes is faulty, failover behavior and network resource preemption behavior of other communication nodes at the same level are triggered. Therefore, how to detect a communication node fault becomes an important problem to be solved.
In a GPRS (General Packet Radio Service, general packet radio service) or UMTS (Universal Mobile Telecommunications System, universal mobile telecommunications system) network, a fault state of a peer GGSN is detected through a Hello message between GGSNs (Gateway GPRS Support Node, gateway GPRS support node); if no response packet is received within a specified duration, a transmitting GGSN thinks that the peer GGSN is faulty, thereby triggering service switchover and network resource preemption behavior.
In a process of implementing the present invention, the inventor finds that the prior art has at least the following problem:
According to an existing fault detection method, if a communication link between two GGSNs is interrupted, the two GGSNs both think that a peer end is faulty; but actually, the two GGSNs may both be normal. Therefore, in the existing fault detection method, a misjudgment may exist.